Altınbaş University Institutional Repository
Not a member yet
5805 research outputs found
Sort by
A multi-branched hybrid perceptron network for DDoS attack detection using dynamic feature adaptation and multi-instance learning
Get PDFThe increasing sophistication and frequency of Distributed Denial of Service (DDoS) attacks necessitate advanced detection systems. These attacks leave networks vulnerable to disruptions, resource overload, security breaches, and financial losses. Conventional detection systems suffer from high false positive rates, lower flexibility, and an inability to adapt dynamically to trending attack patterns. To address these limitations, our proposed work introduces a novel approach to tackling these challenges by merging a multi-branched hybrid perceptron network with dynamic feature adaptation and multi-instance learning. Our methodology features three key innovations: (1) Multi-Branched Hybrid Perceptron architecture, (2) Dynamic Feature Adaption, and (3) Dynamic Attention-Weighted Feature Fusion to improve feature representation and merging process. The proposed study was validated on three testing datasets: (1) UNSW-NB15, (2) CIC-IDS 2017, and (3) CIC-IDS 2018, and the results were compared with various state-of-the-approaches. The experimental results show that our model significantly outperforms existing methods. On UNSW-NB15, the model achieves an accuracy of 96.02% with a precision of 0.965, a recall of 0.963, and an F1-score of 0.9645. For CIC-IDS 2017, it reaches a near-perfect accuracy of 99.99% with all metrics at 1.00. On CIC-IDS 2018, the model performs with an accuracy of 99.96% and perfect precision, recall, and F1-scores of 1.00. Time complexity analysis shows that while the proposed intrusion detection framework takes 21.6 seconds on CIC-IDS 2017, 30.0 seconds on CSE-CIC-IDS2018, and 15.5 seconds on UNSW-NB15, it remains competitive with high performance. Despite its higher time complexity on UNSW-NB15, MHHPN provides superior detection capabilities, making it practical for real-time use in complicated and extensive networks
Optimizing hybrid energy systems: employing smart monitoring networks with IoT integration
No full textFor effective and sustainable energy solutions, this paper investigates at a hybrid energy system that uses Internet of Things technologies. Three components make up the system: PV system, an AC turbine voltage source, DC batteries, and ThingSpeak, an IoT analytics program hosted on the cloud that allows for real-time monitoring. In order to optimize performance, the study emphasizes the significance of hybrid energy systems as well as the requirement for improved analytics and monitoring. A detailed description of the experimental setup is given, including the measured parameters. The results demonstrate the hybrid system's adaptability to various loads and input voltages. fluctuations in DC battery and turbine voltages cause fluctuations in AC power and AC current, while the AC voltage stays constant at 224.4 V. For efficient analytics and monitoring, the hybrid energy system must include ThingSpeak. The study emphasizes how the system can adjust to changing conditions and how Internet of Things technologies may assist hybrid energy systems grow more powerful. This research adds to the current discourse on the integration of IoT data for enhanced system optimization and sustainable energy solutions
Developing a maintenance management system for water treatment plants
No full textThis research focuses on the development of a maintenance management system for potable water liquefaction units, a crucial aspect of infrastructure. The study explores the maintenance of drinking water liquefaction units, analyzing methods, frequency, and duration of maintenance activities. The goal is to create a modern, contextually fitting system for eight water liquefaction units. The existing system has several deficiencies, including inadequate documentation and reporting, and an absence of an evaluation system. This leads to an unstructured approach to planning, scheduling, and monitoring, lacking standardized controls for maintenance performance. The limited budget allocated to unit maintenance results in neglect of less critical tasks, leading to delays in completing essential activities. To address these issues, the researcher proposes multiple subsidiary systems, including a planning subsystem to optimize maintenance costs, a staff allocation subsystem based on unit capacity, a budgeting system, and a reporting subsystem for improved documentation. A project management program (Microsoft Project) is for leveraging its capabilities
Biodegradable polysaccharide aerogels based on tragacanth and alginate as novel drug delivery systems
No full textTragacanth, an anionic polysaccharide, is a natural material widely investigated for the synthesis of aerogels as drug delivery vehicles. Its biocompatibility, biodegradability, and affordability are all key features for its use in pharmaceutical applications. In this study, tragacanth and tragacanth alginate composite aerogels were prepared using the sol-gel technique followed by supercritical drying. Paracetamol was selected as a model drug for drug loading and release studies owing to its high solubility in ethanol and low solubility in supercritical carbon dioxide. The paracetamol loading into the aerogel pores was confirmed by infrared spectroscopy (IR) and x-ray diffraction (XRD) spectra of the resulting samples. Scanning electron microscopy (SEM) images showed that all aerogels were porous with a macroporous-mesoporous network. Due to the high porosity of the prepared aerogels, a loading of 99 wt% (mg drug/mg aerogel) for tragacanth and 114 wt% (mg drug/mg aerogel) for composite aerogels was achieved. Moreover, the release rate of the drug could be modified by manipulating the aerogel composition. Tragacanth aerogels had a faster release rate, while the addition of alginate prolonged the release rate of the model drug. Various empirical release models were investigated and the release rate was found to follow the Korsmeyer-Peppas (Power Law) model suggesting a diffusion-based release kinetics. Based on the results, the feasibility of utilizing tragacanth for the preparation of drug-loaded aerogels was shown.
Tragacanth is a suitable polysaccharide for the synthesis of aerogels.Macropores/mesopores of all synthesized aerogels are suitable for drug loading.The presence of alginate retards the release of the model drug up to 20% at each time point.The release kinetics of the synthesized aerogels follow the Korsmeyer-Peppas model
Is-OWC system using ofdm with hybrid MDM-PDM incorporating DWDM toward tpbs date rate based 5G (LEO-GEO) satellite communication
No full textLong-distance transmission to handle Geostationary Earth Orbit (GEO)-based transmission
and high data rates and enhancing the scalability of Pointing error (Pe) for Tx and Rx
antennas are satellite communication difficulties. This thesis develops a new Intersatellite
Optical Wireless Communication (Is-OWC) system employing hybrid division methods of
Mode Division Multiplexing (MDM) with two Hermite-Gaussian modes (HG00 and HG01)
and Polarization Division Multiplexing (PDM) with two polarizations. 5.12 Tbps, 48,000
km transmission, 8-channel system Results demonstrate reliable system performance over
the planned distance and data rate, with an average log BER of (HG00 = -5.02 and HG01 =
-5.32) for p1 and (HG00 = -4.83 and HG01 = -5.16) for p2 at 48,000 km, beyond the BER
FEC threshold of -2.42. Distance-OSNR inverse relation the recommended system's OSNR
tolerance was < 50.5 dB with Tx and Rx Pe of 1 mrad. When transmission distance rises,
HG00's EVM parameter is bigger than HG01's due to maximum mode power. However,
received power and distance implies a considerable receiver sensitivity at a Pe of 1
milliradian for both the Tx/Rx. A two-coding scheme DP-DQPSK system was built,
examined, and compared to the specified system capacity to test the revolutionary system's
reliability. OSNR varied by 3 dB for Rx Pe up to 2 mrad and 5 dB for higher pe values due
to significant antenna misalignment. Up to 40,000 km, the technique works. Distance
enabled Tx Pe up to 4 mrad. As Tx/Rx aperture diameters grow, geometrical losses and
attenuation are eliminated, lowering expected channel log BER. The recommended method discovered that 5 cm Tx diameter operated well up to 40,000 km, while 10 cm may enable
specific channels reach the needed distance. 15 cm diameter works. 20 cm may increase
system performance, but it needs a Tx/Rx diameter trade-off to minimize transmission
losses. The recommended method transfers data up to 30,000 km with a 10 cm Rx diameter.
20 cm diameter transmission ranges to 40,000 km. 30 cm are appropriate for 48,000 km
transmission. Thus, increasing the Rx aperture diameter above the Tx aperture diameter
reduces geometrical losses and attenuation, which impact antenna gain
Inventory optimization using predictive analytics and machine learning
No full textInventory management is on the verge of a radical transformation with the advent of
advanced predictive analytics and machine learning. The primary goal of the project is to
improve inventory techniques in a variety of business environments using such technologies.
For this purpose, research was conducted to refer to the changes in literature focusing on
inventory management types from traditional to current ones, relying on the analysis of data
processing and interpretation approaches. Thus, this analysis showcased a transition towards
more dynamic, adaptive and predictive inventory control methodologies on one side.
Furthermore, a research design which implies methods like data collection and preparation
and feature engineering and multiple machine learning model’s implementation was also
conducted. Finally, a group of case studies illustrating effective results after applying
predictive analytics was considered to support the claims presented in the opening section.
Despite the potential importance of the problem discovered in the study, several key
limitations were acknowledged. The most crucial identified issues included data quality
concerns, multi-disciplinary aspects, and the problem of ethics. At the same time, the
expected effects of predictive analysis applicability were found promising and, on their way,
to realising only through the study and research.Envanter yönetimi, gelişmiş tahmine dayalı analitik ve makine öğreniminin ortaya
çıkmasıyla radikal bir dönüşümün eşiğinde. Projenin temel amacı, bu tür teknolojileri
kullanarak çeşitli iş ortamlarında envanter tekniklerini geliştirmektir. Bu amaçla, veri işleme
ve yorumlama yaklaşımlarının analizine dayanarak, gelenekselden günümüze envanter
yönetimi türlerine odaklanan literatürdeki değişikliklere atıfta bulunmak için araştırmalar
yapılmıştır. Bu nedenle, bu analiz bir tarafta daha dinamik, uyarlanabilir ve tahmine dayalı
envanter kontrol metodolojilerine doğru bir geçiş sergilemiştir. Ayrıca, veri toplama ve
hazırlama ve özellik mühendisliği ve çoklu makine öğrenimi modelinin uygulanması gibi
yöntemleri ima eden bir araştırma tasarımı da yapılmıştır. Son olarak, tahmine dayalı analitik
uygulandıktan sonra etkili sonuçları gösteren bir grup vaka çalışmasının, açılış bölümünde
sunulan iddiaları desteklediği düşünülmüştür. Çalışmada keşfedilen sorunun potansiyel
önemine rağmen, birkaç önemli sınırlama kabul edildi. Tanımlanan en önemli konular
arasında veri kalitesi endişeleri, çok disiplinli yönler ve etik sorunu yer alıyordu. Aynı
zamanda, tahmine dayalı analizin uygulanabilirliğinin beklenen etkileri umut verici bulundu
ve yalnızca çalışma ve araştırma yoluyla gerçekleşme yolunda ilerledi
A hybrid ensemble learning approach for efficient diabetic retinopathy prediction and classification using machine learning and deep learning techniques
No full textAI is a crucial tool in early detection and classification of diabetic retinopathy, which is a leading cause of visual impairment globally. Transfer Learning (TL) was used to improve the accuracy of predictions and classifications within training datasets, surpassing existing methodologies. The study provides comprehensive insights into current databases, screening programs, performance evaluation metrics, relevant biomarkers, and challenges encountered in ophthalmology. The findings underscore the potential of AI-based approaches in enhancing diagnostic precision and offer a promising direction for future studies. The paper concludes by delineating opportunities for further research and development in integrating AI advancements in the field. Conclusion: The findings underscore the efficacy of Transfer Learning in significantly improving the accuracy of diabetic retinopathy image predictions. This research highlights the potential of AI-based approaches in enhancing diagnostic precision and offers a promising direction for future studies. The paper concludes by delineating opportunities for further research and development, emphasizing the continued integration of advanced AI methodologies in ophthalmology to advance diabetic retinopathy detection and management
Numerical and experimental analysis of deformation in cantilever and anchorage sheet piles
No full textThis study systematically examines the behavior of cantilever and anchorage sheet piles influenced by a range of variables, namely dredge level, water table level, anchor interanchor distance, anchors’ number, and the distance from the ground surface. Utilizing a single sheet pile material, the investigation highlights the deformations resulting from these variables, in addition to assessing the effects of different soil types and degrees of saturation. The primary objective centers on the computation of bending moments arising from these factor variations under a constant load, using the finite element theory-based software, Plaxis 2D Connect Edition V20. Sheet pile behavior could be evaluated using this tool by investigating driven depth, maximum bending moments, and horizontal displacements in both anchorage and cantilever sheet piles. To enhance the reliability and realism of findings, outcomes are compared with results from the Pro Sheet program. This comprehensive evaluation furnishes valuable insights into sheet pile deformations under diverse conditions, contributing to the existing body of knowledge and facilitating more robust engineering practices
The impact of incorporating five different boron materials into a dental composite on its mechanical properties
No full textThe aim of this study is to comparatively investigate changes in mechanical properties by adding five different types of boron derivatives to a microhybrid dental composite structure. In this study, which evaluated upper and lower surface microhardness (VHN), roughness (Ra), and color stability (Delta E), a total of 126 discs were used (n = 7; per each subgroup). All boron derivatives were added to the dental composite structure in equal proportions in a dark room to create experimental composites (5% w/w). To enable comparison, a default composite without the addition of any boron derivative served as the negative control group. Before measurements, all samples were incubated at 37 degree celsius for 24 h. For surface microhardness, roughness, and color stability, all experimental groups were statistically significant within themselves (p < 0.001). The group with borax pentahydrate exhibited the highest VHN value on the upper surface, while the lowest value was observed in the group with etifert. It was found that all experimental groups showed a decrease in lower surface microhardness compared to the control group (p < 0.001). Although the eticol-ceramic-added group had the lowest roughness values, this group also exhibited significantly higher increment E values compared to the other groups. Surface roughness showed a negative correlation with increment E and upper surface microhardness values for all experimental groups (respectively p = 0.038; r = -0.185/p = 0.006; r = -0.245). To sum up, the addition of boron derivatives to composites, except for etifert, increased upper surface microhardness values; however, except for eticol-ceramic, surface roughness values also increased. Nevertheless, the addition of boron derivatives, except for eticol-ceramic, ensures color stabilization
Architectural appraisal and current condition assessment of Ottoman minarets in İstanbul
No full textMinarets are considered one of the most prominent monuments that characterize Islamic
architecture, which contributed to its prosperity. According to historical studies, the idea of
minarets arose before Islam, and in the Islamic period, minarets spread, and their shapes,
design methods, and methods of construction developed from one era to another, especially
in the Seljuk, Fatimid, Mamluk, and Ottoman eras. The most aesthetic and most famous
minarets, they are distinguished by their pencil shape, graceful body, and high height. Due
to the influence of time, weather factors, climatic effects, and natural disasters such as
earthquakes, these minarets are exposed to great damage, represented by biological and
structural damage.
Therefore, this study is concerned with evaluating the Ottoman minarets in the city of
Istanbul, being the city of minarets. A reconnaissance study was conducted for ten famous
minarets in the city of Istanbul, and their damage was assessed from a structural and
biological standpoint. Remedial solutions were also proposed according to the results of the
exploratory study to evaluate the current conditions of the minarets